Imagine universe without any humans (or any other sentient beings). From my perspective, talking about “morality” in such universe simply does not make sense, this word does not apply to anything that exists there.
Depends on who does the talking. Why would presence of something in the universe influence the methodology of judging it (“can’t judge it”), rather than the result of a judgement (“it’s worthless”/”it has no moral relevance”)? (Sounds like corrigibility, a morality that is not in closed form and depends on environment.)
Right. So for that to make sense, the things being judged are not the universe as a whole (or itself), but some sort of parts/aspects abstracted from it, objects of a different kind that are only relevant by somehow relating to it, perhaps “embedded” in it.
This is harder to set up as a guide to decision making, because consequences of actions or decisions are not as isolated from the rest of the universe, but I guess scoped consequentialism (goodhart/boundaries, mild optimization) would want to make some sense of this. Also, updateless decisions isolate abstractions of ignorance about current/future observations.
Depends on who does the talking. Why would presence of something in the universe influence the methodology of judging it (“can’t judge it”), rather than the result of a judgement (“it’s worthless”/”it has no moral relevance”)? (Sounds like corrigibility, a morality that is not in closed form and depends on environment.)
The absence of something can easily preclude the possibility of judging it.
Right. So for that to make sense, the things being judged are not the universe as a whole (or itself), but some sort of parts/aspects abstracted from it, objects of a different kind that are only relevant by somehow relating to it, perhaps “embedded” in it.
This is harder to set up as a guide to decision making, because consequences of actions or decisions are not as isolated from the rest of the universe, but I guess scoped consequentialism (goodhart/boundaries, mild optimization) would want to make some sense of this. Also, updateless decisions isolate abstractions of ignorance about current/future observations.